As the UK government announce plans to expand testing rollout, Stephen Reicher and colleagues argue that testing without support for self-isolation won’t work
For many months, there has been concern at the low levels of adherence among those asked to self-isolate because they have symptoms or a positive test for the coronavirus, or are contacts of a positive case. Indeed, data collected as part of the CORSAIR study between March 2020 and January 2021 suggested that only 42.5% of UK respondents were fully adhering to the requirements when symptomatic.  In its latest wave of data, collected at the end of January, adherence had risen slightly to 51.8%, but that still means only half of those who should are self-isolating.
Given evidence that those from lower socio-economic groups and with fewer resources are least likely to self-isolate—based on international comparisons showing the effectiveness of comprehensive support packages, and from survey studies showing greatly increased intention to self-isolate when compensation is available—many have called for such support packages in the UK in order to improve levels of adherence. [2-4]
In response, the UK Government introduced a £500 self-isolation grant for those on benefits in September 2020. This was enhanced by a discretionary fund of £20 million per month announced at the March 2021 budget and then, later in March, further enhanced by an additional £12.9 million per month plus £3.2 million per month for a medicine delivery service to those having to stay at home. [5-7]
This response has been criticised as inadequate along multiple dimensions. Firstly, only about one in eight of the workforce are eligible for the £500 compensation. Of those who apply only about 30% succeed, and of those who apply for the discretionary scheme only 20% succeed (the figure in some regions, such as the North West, falling as low as 10%).  Secondly, the financial sums available are inadequate, since £500 over 10 days is less than the minimum wage and not enough for most people to be able to forego their income from working. Other countries pay a generous percentage of income, rising in some countries such as Finland where 100% compensation for pay is available.  Thirdly, the scheme needs to be far more comprehensive and address all the needs that people have: these include accommodation (how can one self-isolate in a crowded multi-generational household?), caring needs (what do you do if you have children or elderly relatives to look after?), and even walking the dog. Wrap-around packages are available not only in affluent places like Taiwan and New York (where they will indeed walk your dog), but also in far poorer locations such as Vietnam, Uruguay, Cuba, and the Indian state of Kerala. [10-13]
All in all, many have argued that UK support for self-isolation remains highly inadequate and constitutes possibly the biggest flaw in our pandemic response.  After all, testing and contact tracing on their own are of limited use. Both are means to identify those who may be infected and take them out of the chain of transmission. For want of funding self-isolation with a small fraction of the £37 billion spent on privatised test and trace, we undermine the whole system. It makes sense neither in public health nor economic terms.
But a week ago, all this seemed to change. ONS figures released on 26 March 2021 suggested that, in the period 1 February to 13 February, 86% of people fully self-isolated.  In the period of 1 March to 6 March, this figure increased still further to 90%.  These findings were welcomed by Matt Hancock who said: “I am delighted that today’s statistics show the vast majority of people in this country are doing the right thing and following the rules to help protect the NHS and save lives”.  The problem seemed to have gone away. If nearly everyone is now self-isolating, then the existing support is doing its job and there is no need to allocate more money to increase it.
So why such a dramatic discrepancy between the ONS figures and previous findings—notably the CORSAIR study? What is going on? There are many points that could be made: ONS themselves acknowledge that these are “experimental data” and that the methodology is still in development; the ONS had a response rate of just 19%, the CORSAIR survey used quota sampling; the ONS contacted people by phone, CORSAIR was conducted online by market research companies.
But the key difference is this: the ONS study is limited to people who have engaged with the Test and Trace system and are known to it. They are those who have already adhered in terms of getting a test and providing contacts and moreover are among the one fifth who also adhered in terms of answering the ONS survey. What the ONS survey finds, then, is that the most adherent among the adherent also adhere when it comes to self-isolation. By contrast the CORSAIR study involves a general population sample of anyone with covid-19 symptoms. It shows that when you widen your net to include people, whether or not they have had contact with the Test and Trace system, adherence is much more of a problem.
To put it slightly differently, what the difference between the two studies shows is that the problem is actually rather wider than has hitherto been emphasised. It isn’t just about self-isolating, it is to do with engaging with the Test and Trace system in the first place. Indeed the biggest problem is the remarkably low numbers of people who even get a test (18.8% overall according to the CORSAIR findings, rising to 22.2% in the latest wave of data collection). This might seem to indicate that we have laid too much emphasis on getting positive cases and their contacts to stay at home. However, arguably it points to the fact that self-isolation is even more important than has been argued to date.
That is because there is evidence that a key reason why people don’t engage with the testing system at any point has to do with the costs of self-isolation—and the penalties for breaking self-isolation. For instance, during mass testing in Liverpool in November 2020, people living in the most deprived fifth of areas were half as likely as those in the most affluent fifth to get a test. The Interim Evaluation report points to concerns about losing income, or even losing one’s job, as a driver of this difference. As the report puts it “these concerns could be expressed as a feeling of being overwhelmed by the implications of a positive test—‘I can’t face it’” (p.24).
By the same token, even if you decided to get tested, why would you provide the names of your family and friends as contacts if the consequences for them were equally severe? These consequences became even more severe with the introduction in England in September 2020 of £10,000 fines for failing to self-isolate. This then left poorer groups in a bind: not being able to afford to self-isolate and not being able to afford to not self-isolate. The only way out of this bind was to avoid testing and avoid providing contacts to put friends and family in the same bind. As a paper modelling effects of fines by SAGE scientists put it “Legal enforcement of self-isolation can create trade-offs by dissuading individuals from self-reporting.” They concluded: “policies such as fines, and police enforcement of self-isolation will have either little benefit or a negative effect.” 
If the importance of self-isolation is actually greater than we thought before, it is equally true that the need to improve support for self-isolation is more urgent as a policy priority. What is more, this urgency will become yet more acute as covid restrictions are eased over the coming weeks and months and the potential for infected individuals to mix with others and transmit the virus becomes all the greater.
We need to get to a point where we are able to deal with infection outbreaks through a targeted policy of rapid testing, forwards and backwards testing, and of taking infected people out of community circulation. We cannot do that without better support for isolation. So, without support for isolation, talk of a “road map” that is irreversible and of never returning to lockdown is precisely that. Just talk.
Stephen Reicher, School of Psychology and Neuroscience, University of St. Andrews.
John Drury, School of Psychology, University of Sussex.
Susan Michie, Division of Psychology and Language Sciences, University College London.
Competing interests: All authors participate in the UK’s Scientific Advisory Group for Emergencies and/or its subgroups and in Independent SAGE. All are writing in a personal capacity. SR participates in the advisory group to the Scottish chief medical officer.
Funding: The work of SR and JD on this paper was supported by a grant from the ESRC (reference number ES/V005383/1). SM is a co-Investigator of the COVID-19 Rapid Survey of Adherence to Interventions and Responses [CORSAIR] study funded by NIHR Evaluation, Trials and Studies Coordinating Centre.
The views expressed are those of the author(s).
1] Smith, S.E., Potts, H.W.W., Amlot, R., Fear, N.T., Michie, S. & Rubin, G.J. (2020) Adherence to the test, trace and isolate system: results from a time series of 21 nationally representative surveys in the UK (the COVID-19 Rapid Survey of Adherence to Interventions and Responses [CORSAIR] study)
2] Atchison, C., Bowman, L. R., Vrinten, C., Redd, R., Pristerà, P., Eaton, J., & Ward, H. (2021). Early perceptions and behavioural responses during the COVID-19 pandemic: a cross-sectional survey of UK adults. BMJ open, 11(1), e043577
3] Patel, J., Fernandes, G., & Sridhar, D. (2021). How can we improve self-isolation and quarantine for covid-19?. British Medical Journal 372:n265
4] Bodas, M., & Peleg, K. (2020). Self-Isolation Compliance In The COVID-19 Era Influenced By Compensation: Findings from a recent survey in Israel: public attitudes toward the COVID-19 outbreak and self-isolation: a cross sectional study of the adult population of Israel. Health Affairs, 39(6), 936-941
10] Patel, J., Fernandes, G. & Anchuri, Kavya (2021) Self-isolation-related support, monitoring and adherence: A scoping review of international approaches. https://static1.squarespace.com/static/56ebbd6827d4bdff1f7e7ae1/t/600ff28547f88841394166bb/1611657863707/Covid+Isolation+Review+GHGP+20012021.pdf
18] Liverpool Covid-19 Commuity Testing Pilot: Interim Evaluation Report (2020)
19] Lucas, T., Davis, E., Ayabina, D., Borlase, A., Crellen, T., Pi, L., & Hollingsworth, T. D. (2020). Engagement and adherence trade-offs for SARS-CoV-2 contact tracing. Philosophical Transactions of The Royal Society B Biological Sciences.