Of prime importance to governments and the public is how and when lockdown can be lifted. Attempts to infer from previous pandemics (eg, H1N1, SARS) are fraught with challenges given continuing uncertainties about parameters as fundamental as the rate of transmission (R or more accurately R_t, the effective reproduction number) and case fatality rate [6]. Both figures are very difficult to estimate reliably given the lack of a true denominator of the number of cases in the UK [7-9].

To bring clarity for future decisions, we have identified and mapped the range of approaches that are either actively being explored or have been implemented internationally (Table 1). The objectives of these approaches are either to: i) promote natural herd immunity; ii) release lockdown measures once a treatment or vaccine become available; iii) attempt to completely eliminate the virus; or iv) find ways of relaxing lockdown whilst still containing virus transmission through a ‘test, trace, isolate’ strategy. Such measures would need to continue until an effective treatment is available or herd immunity (whether naturally or through mass vaccination) develops. We briefly consider each of these approaches in turn.

.  Photo: Liverpool Street Station, London. By Ben Garratt via Unsplash.

The UK, Dutch and Swedish governments initially pursued the idea of promoting herd immunity, but the UK and Dutch have subsequently abandoned this strategy in light of mounting public concerns about the need to infect 60%-80% of the population and revised modelling, which indicated that health systems would be over-whelmed and this could not be pursued as a feasible public health policy [21]. Sweden is in contrast still broadly pursuing this approach with only minimal enforced restrictions on physical movements. Its case is however somewhat atypical as a large proportion of its population ( ~ 40%) live in single person households, it has a highly skilled workforce which enables many people to work from home, and the high levels of education have led to general compliance with messaging about maintaining physical distancing leading to relatively low levels of infection (estimated at ~ 7%) [22]. It is as yet an unproven assumption that long-term immunity is possible despite some evidence of potential reinfections and short-lasting immunity from other coronaviruses [23]. Of note, Sweden’s approach has resulted in significantly more deaths than comparable Nordic countries [24].

Considering the second option, optimistic estimates suggest that a vaccine will not be available for at least 12-18 months and even then it is unknown how safe, effective, available or accessible such a vaccine would be [25]. Given the decimating health, economic and social impacts of lockdowns, it is practically inconceivable that widespread lockdowns can continue even beyond the next 4-6 months. That said, this cannot be a license to recklessly lift lockdown restrictions too early as, for example, the Mayor of Las Vegas was considering [26].

A ‘test, trace and isolate’ strategy is the cornerstone of managing epidemics, and has been repeatedly emphasized by the WHO. Countries such as China (in particular Hubei Province), Hong Kong, South Korea, Taiwan and Vietnam have pursued such approaches achieving substantial control of the disease. Importantly many of these countries had previous experiences of dealing with SARS and MERS, which enabled them to understand the magnitude of the potential challenge very early on and mount effective nationally coordinated multi-sectoral responses. In contrast, the UK and many other Western nations were delayed in instituting their responses, which allowed the numbers of cases to rise very rapidly. Also of relevance in this respect is that these Asian countries there has been active surveillance and enforced isolation of suspected and confirmed cases. Such approaches are far less socially acceptable in Western European contexts. The success of such approaches is also highly dependent on the scale of testing and contact tracing achievable and the efficiency with which test results are turned around. There are therefore a number of substantial challenges that need to be overcome to effectively implement ‘test, trace and isolate’ strategy in the context of the UK [27].

In the UK, the window of opportunity has passed for quick elimination of the virus, given active community transmission as well as the absence of surveillance, testing and tracing capacity. If sufficient political will is found and public health infrastructure is created, it could become a route in the medium- to longer-term. This would however require public acceptance of some of the privacy infringements that contact tracing and an active quarantining approach rely on, which have been employed in a number of other countries. For example, it is estimated that 10 000 new symptomatic cases per day would require 140 000-390 000 contacts to be quarantined/d [28]. Given these requirements, it is currently not feasible to pursue elimination as this would involve maintenance of tight lockdown, vast armies of contact tracers as well as a huge increase in test kits to allow for population testing and extensive quarantining capacity or monitoring of cases.

The last option is we believe the most credible for the UK in the short-term. In essence, this involves finding ways of living with the virus by gradually relaxing some of the lockdown measures whilst simultaneously pursuing aggressive development of active surveillance, testing, tracing and quarantining capacity. These are interim interventions to contain the spread of the disease, keep daily new cases low and minimise its impact until a vaccine, effective treatment, or more data on immunity becomes available. This is likely to involve the careful, gradual opening of segments of society – for example, beginning with predominantly outdoor workers who can maintain physical distancing and/or geographical areas of the country where there is the prospect of eliminating the disease (eg, Scotland’s island communities). This could be followed by a careful phased re-opening of schools with reduced class sizes, desks at 2m distances, face masks for older pupils, frequent hand washing; and the gradual re-opening of non-essential businesses whilst maintaining physical distancing measures. We envisage certain sectors such as bars and restaurants will not open until much later. British Columbia is, for example, considering easing lockdown measures guided by modelling that predicts that increasing social interactions from the current 30% to 40%-60% would not result in a second outbreak [29].

The idea of ‘bubbles’ has been floated whereby each household has interactions with a limited number of others. Recent research has suggested that limited, repeat contacts within the same group may be an effective way of allowing limited socialisation whilst still keeping cases down [30]. This approach does however have the potential to trigger family discord and also has the challenge of being very difficult to monitor and enforce.

Any lockdown relaxation measures need to be considered alongside provision of personal protective equipment (PPE) for those most at risk of acquiring or transmitting the virus and continuing to shield the most vulnerable members of society, including those in care homes. The impact of such changes in policy will need to be continuously monitored using real-time, multi-modal data with a willingness to (temporarily) tighten controls when necessary. Germany recently began to cautiously ease its lockdown measures, a federal rather than national decision, and has very quickly seen a rise in R_t from 0.7 to 1, leading to considerations to retighten the restrictions.

What is clear is that there is no straightforward way out and that moving forward will be a delicate balance involving difficult trade-offs. Whichever approach is taken, transparency and clear communication with the public is essential as well as recognition that apprehension might exist to re-engaging with aspects of public life [31].