Full working paper: evolving lockdown

April 25, 2020 • Reading time 16 minutes

All of our analysis on the impact of Covid-19 on the NHS is shared here. For further information please contact George on 07980804956 or [email protected]


This is a working paper on how to evolve lockdown. A summary is available here. We welcome feedback, which you can provide here, or via emailing [email protected].


Published 27 April 2020


1. Introduction: evolving lockdown in the UK

2. Objective: the health and wealth of nations

3. Protective measures

4. Locations of potential risk

5. Reopener innovations

6. Preparing for the next pandemic

1. Introduction – evolving lockdown in the UK

Lockdown is the new pandemic. People are dying because of it; and the economy is taking a hit, which will lead to more suffering for more people in the long run.

It is not possible to test everyone and to track the disease at the level needed to end the uncertainty of where the disease is at any given point in time. This challenge is before considering the accuracy of tests.

Nor is it possible to go down the discredited “herd immunity” route or anything similar – unless new evidence becomes available that dramatically reduces the risk to individuals (e.g. accurately identified individuals that need to be shielded).

These challenges mean that ending the lockdown will be a series of smaller steps, many of which can start now. There will also need to be a range of measures in place to ensure that as the lockdown is released it is safe and there are protocols in place to allow quick and corrective actions if anything is not working as hoped.

There is a genuine risk that delaying these steps in the hope of a vaccine or a “test and trace” panacea will result in delay and consequent non-compliance with current social distancing. A delay could contribute to a second Covid-19 peak, further lockdown and more uncertainty.

2. Objective – the health and wealth of nations

The aim of ending the lockdown should be to allow the economy to continue as much as possible without coercing anyone into harm’s way or putting the health of the nation more generally at risk – short or long term.

This means:

  • Putting in place protective measures, such as requirements for mask-wearing, temperature checks and continued shielding of the vulnerable
  • Recognising that removing the lockdown will necessarily create some population health risk, so taking protective measures to reduce the level of risk where possible pragmatically
  • The removal of lockdown will need enforcement and rapid assessment of what may or may not be working so that this is flexible to demands there should be space to allow “reopener” innovations
  • Recognising the challenges involved to date and going forward (this and the next pandemic) it is also essential to consider the approach needed to insure against a possible resurgence

3. Protective measures

While a lot is still unknown about the virus that spreads the Covid-19 disease, there are some basics, such as it spreads through respiratory droplets that can survive on surfaces for several days.

This method of spreading means we can say with some confidence mass gatherings, particularly those involving close contact, or a lot of singing or shouting should not be permitted. Other activities, such as eating at restaurants or drinking at bars, are also high risk as involve people from different households mixing in a single location – see “locations of potential risk” section.

It is unclear exactly how these “social” activities can revert to pre-Covid-19 without the presence of herd immunity or vaccine. However, there are perhaps process and procedure innovations that can enable some businesses to return to a new normal – see “reopener innovation” section.

But it is possible to consider some protective measures that can be taken now at relatively low cost (i.e. compared to high mortality and economic collapse). Phasing will help reduce the risks involved. There should also be monitoring and quick evaluations. These include:

a) Masks

b) Temperature checks

c) Frequent disinfection and cleaning

d) Contact tracing

e) Shielding of the vulnerable

3a. Masks

There is growing agreement that people wearing non-surgical masks benefits the broader community even if the wearer remains unprotected – these can be made of cloth. This is because they stop people, who may be asymptomatic, from unknowingly spreading the disease. For example, the droplets from a sneeze go further when they are not behind a cloth mask.

These would not need to be worn everywhere or all the time, but in certain areas where the risk of spreading the disease is higher (e.g. on public transport).

3b. Temperature checks

The ability to test enough people with antibody tests is limited and slow. Given 71% of laboratory-confirmed cases of Covid-19 also had fevers, temperature checks are the right solution for high volume and low-cost testing that can help pinpoint potential cases quickly and routinely.

There would need to be careful thinking where to use these and how not to identify too many people (e.g. joggers). Positive tests would also be an indicator for greater vigilance and, ideally in some cases, antibody testing.

3c. Frequent disinfection and cleaning

There has been significant focus on washing hands and not touching faces, although there has been less visible cleaning of high-risk surfaces (lift buttons) and locations (train stations and supermarkets).

Cleaning standards should be set and cover the level of disinfection (e.g. what substances are used to clean) and frequency (e.g. every two hours).

3d. Contact tracing

A lot of hope is being placed on contact tracing technology to help end the lockdown when combined with antibody testing. The concept is that technology identifies, anonymously, people that you recently contacted. If one of these people if identified as having the disease, then an alert is sent to others that were in contact with the infected individual. These people are then required to isolate for a necessary period.

This technology may well help, although it is unproven and there are situations where it seems less likely to work (e.g. the Underground where there are too many people too close for a short time).

A complimentary piece of technology might be something that alerts the user to the relative number of “localised” cases and gives an indication of safety – in China areas with fewer cases get “green” and ones with more cases “red”.

With both of the “apps”, they will only work if the information or nudge they provide changes individual and group behaviour.

3e. Shielding of the vulnerable

There is currently an identified population that is considered high risk from Covid-19 and shielded. As new data become available, it may be possible to reduce or extend these shielding protections to more or fewer people.

Maintaining the shielded status, until there is a vaccine or some form of protection, will prevent mortality and demands on the NHS rising substantially.

4. Locations of potential risk

Restarting economic activities required opening locations for use. Most of these create a risk. These include:

a. Health and care centres – hospitals, GP surgeries, care homes, etc

b. Mass transit systems – underground, trains, buses, etc

c. Places of education – schools, colleges, etc

d. Places of work – factories, offices, etc

e. Society in general – socialising, shopping, etc

These areas unavoidably involve (often large) groups of people in proximity, so could allow the spread of the disease at the rates seen in March if reopened without any change.

4a. Health and care centres

By their nature, hospitals and GP surgeries are where people that are sick go. They are unavoidable, although many people appear seem to have chosen to avoid hospitals – presumably based on the risk of infection and ending up behind closed doors in the last moments of life.

Hospitals are mainly good when it comes to infection control protocols, although based on recent personal experience there are few ‘front of house’ measures being followed aside from reduced access, door security and lots more hand sanitizer. There is a lack of social distancing or mask-wearing – this makes the communal areas of hospitals feel more normal than anywhere else in London at present.

The challenge is that the focus has been to stop people (where possible) coming in and spreading disease. This has precipitated a shift to video consultations and no family visiting, but less has been done at the front of house to reduce risk consistently. Measures could include things like:

  • Everyone (staff, public, etc.) should be required to wear face masks while walking around a hospital.
  • Directional walkways (left and right-hand sides) and two-meter markers (not always possible) to help people distance.
  • Increased frequency of cleaning (2-hourly if not already) and increased focus on high transmission areas (buttons, handles, etc.). (Note it is unclear how much this is done at present, but it was not visible.)
  • Consider propping open doors that are not automatic, including to bathrooms.
  • These are simple processes that would seem sensible. As with facemasks, the precautionary principle would seem to apply even if there is currently a lack of evidence to support measures.

4b. Mass transit

Transport networks are necessary for other economic activities, so it should be a core part of reopening the economy. On the Underground there were over five million journeys per day – many carriages hold up to 170 people, often in very close proximity.

This section focuses on the London Underground as it is the epitome of mass transit in the UK, so anything that works for the Underground should also work on buses and trains. It may, however, be possible to reduce some of the requirements set out below on other networks if it can be established the measure is not required, and its removal would be safe.

A test for the Underground is what would

  • Establishment of travel permits linked to user accounts – so that rules can be enforced
  • Travel slots to limiting the number of people travelling at a given time
  • Temperature checks at all stations
  • Mandatory mask-wearing, failure to comply results in removal of travel permit

Travel permits

In 2019 there were around 5 million journeys per day on the Underground. Reducing crowding is hard when many people are using the system, but the data suggests this will help reaccelerate the spread of Covid-19.

Travel permits would allow TfL to control the number of people using the system at a given point in time. Removal of these would also encourage people to follow new travel rules, such as wearing a cloth mask.

Travel slots

Allocation of travel slots should be fair – e.g. each Londoner gets three return journey slots per week. More flexible slots could be allocated to key workers.

Some of these slots should be untransferable, but others could be permitted to be transferred. This would allow people not needing to travel to work to transfer their slot to someone else.

The fast-moving nature of the virus and many policies and behavioural changes mean these slots should only be available at any point in time for the new week or two.

Temperature checks

These were covered in the previous section in more detail. In terms of their use on the Underground or other transport networks, these should be deployed at all stations and within large interchange stations (Bank, Waterloo, etc.).

Ideally, temperature checking would also be linked to user profiles for contact tracing if this becomes workable.


These were covered in the previous section in more detail. Given the close contact of people and surfaces, masks (which could be cloth) should be compulsory, and this should be enforced – once enough people wear masks on the Underground, social factors will likely drive enforcement.

Frequent disinfection and cleaning

This was covered in the previous section in more detail. In terms of its use on the Underground, the carriage should be disinfected on the completion of a run (i.e. at the end of the line). Ideally, they would also be cleaned mid-route, although taking people off and on the trains might cause more harm (e.g. if people congregate).

4c. Places of education – schools, colleges, etc

Schools and colleges have been closed since late March. This will start to have an impact on the development of children and the future potential of the economy. Closures also mean many parents are unable to work effectively. For the reasons (and likely more) schools should be a priority for being reopened.

Children are primarily considered low risk, although as potentially asymptomatic carriers of the virus, they pose a threat to older family members.

One way to reduce the risk of reopening schools is phased opening, such as with a two-day week (possibly involving longer hours). This would reduce class sizes, and with cohorting of children could reduce the risk of the disease spread too far too quickly.

Reopening should ideally be combined with temperature checks if these are shown to be useful for detecting the virus in children. There should also be random and frequent (e.g. twice a week) antibody testing of children to establish if there are signs of the disease spreading. On the finding of a positive result, there should be agreed protocols for testing other children in a cohort, isolating those found to be infected, cleaning the premises and reopening as soon as safe to do so.

For high-risk children or those with shielded family members, it may be necessary to consider enhanced homeschooling or, preferably, a high-risk class that is subjected to high levels and frequency of testing (e.g. daily).

4d. Places of work – factories, offices, etc

Getting people back to work is key to reenabling economic activities. Some businesses can be active remotely. But in many places of work, there will be unavoidable contact, so there is a risk.

The primary objective should be to reopen “closed” businesses. These are the ones that are currently not active as they require people in person, not on a video call, such as factories. Where risk can be limited, (e.g. maximum number of people “on the shop floor”) these should be prioritised for some relaxation of social distancing rules.

Given the political pressure likely to come from businesses that are not permitted to “reopen” there should be measures to allow them opening hours (e.g. two days a week). This will unlikely work in the long run, but it will help reduce the pressure for action and likelihood of mistakes.

Reopening business also poses a risk to individuals being coerced back into an “unsafe” environment. These individuals run the risk of being discriminated against if they chose not to travel to work, so some safety measures should be put in place – these could be requirements on businesses or support for individuals.

4e. Society in general – socialising, shops, etc.

There is evidence that isolation is going to be harmful to an individual’s mental health and wellbeing. This is likely to be particularly true for (i) overcrowded households, particularly those with children, (ii) people living on their own, and (iii) people in potentially abusive relationships.

The current ‘household’ rule has worked well due to its simplicity, but it needs to evolve. One option (“reopener innovation” perhaps) would be to permit the creation of virtual households for up to five people in a geographic area. Virtual households should be allowed rights of actual households – walks together, etc.

Building on this concept, communities should be allowed to form to let some socialising. Without this, it is a matter of time before people break the current rules.

Communities could be licenced, but self-governed (i.e. decide membership). There should be templates for how these can be formed and a requirement to record members. Data should be collected to help inform the understanding of the disease and further development of rules.

5. Reopener innovations

Until there is some clarity on when all lockdown measures will end, innovations that help reenable economic activities to happen safely should be permitted. They will support the economy and perhaps stimulate new activities or innovations that were crowded out from the pre-Covid-19 economy. These will be particularly important for businesses that involve people socialising, such as restaurants and bars.

Reopener innovations are less likely to be “technology”, such as contract tracing apps (see “Protective measures” section), but rather process and procedures that create safety by design – not by notification.

These process and procedure innovation need to be “safe” and observed closely at the start. However, if the cost is high (e.g. allowing restaurants to open with only one table per sitting) or potentially creates inequality (e.g. requiring a smartphone app), it should be targeted for removal.

Once proven and sensible processes and protocols are established, these should be turned into rules that can be enforced – perhaps through a licencing regime. Non-compliance could result in a fine. The rules should be relaxed as quickly as possible – the objective is to allow businesses to reopen and contribute to the economy, not to make business hard or fine them.

Example of what innovation could mean for a restaurant

Some restaurants have already found ways to innovate around the lockdown – e.g. refocus on takeaways or food delivery services. But broader innovations may start to become possible and attractive, such as a bookable restaurant with booth-based seating for households that have intelligent scheduling (arrival slots) to minimise transmission risk.

To allow these innovations to happen, there should be mandatory requirements to ensure these innovations are as safe as possible in advance of being tested and can be assessed and controlled if evidence emerges that they are not working.


  • Temperature checks for everyone entering restaurant – staff and guests
  • Everyone should be required to wash and sterilise hands-on entry
  • Doors, including those for the bathrooms, should be automatic or propped open
  • Face masks mandatory for everyone working and should be worn by guests until they are seated
  • Only accept pre-booked tables for households of people, possibly with verification of people being from the same household
  • Pre-booking only possible if all people enable contact tracing app and “commit” to keep it switched on for at least 14 days
  • New regulations for restaurants that want to open to enforce hygiene steps taken and risks minimised

Some of these steps, such as pre-booking or requirement for contact tracing app, might be undesirable from equality perspective – e.g. people without ability access or desire to use technology. These requirements should be relaxed as quickly as possible. (Note: this does not mean data should not still be collected and analysed more than needed.)

6. Planning for the next pandemic

Whatever your views on the Government’s response to this pandemic, there have been gaps in understanding and responsiveness:

  • Ventilators were a big concern before lockdown
  • Oxygen use at one hospital got so high during peak demand the system started to fail and it had to close its A&E
  • PPE remains a contentious topic as frontline hospital staff are sent into potential danger without protection
  • Testing capacity was identified early but remains a challenge

There will likely be other shortcomings of the systems we currently have, so this list will inevitably get longer.

It is not reasonable, nor sensible, to have all these things sorted immediately nor to get them ready in advance of the next pandemic. In essence, even if they were, each pandemic will likely come with a different set of requirements and challenges. This means the key ingredients for future preparedness are:

a. Buying time

b. Developing understanding

c. Responding quickly

6a. Buying time

While each pandemic is different, protective equipment feels like a constant. On the assumption that any epidemic can become a national crisis, this equipment needs to available quickly on mass. The quantity stockpiled should be related to the estimated time for supply chains to start producing and distributing amounts that are required to meet demands. Demands should be calculated based on the potential speed of the spread of diseases (we are all “experts” on this now) and expected usage. This will should be based on a series of potential scenarios.

Government responses will always lag due to the need to be clear and definitive in their action and aversion to creating problems without considering the potential unintended consequences of their actions (e.g. panic buying). This means business and community responses should be enabled as much as possible with some understanding of how Government might react – e.g. at Edge we stopped working in the office or travelling to clients before the requirement to do so as we had a good understanding of the potential risk.

Publicly shared protocols and guidance, perhaps from an independent non-Government department, could be created and available in advance, so that business and communities could build these into their responses.

6b. Developing understanding

Publicly available information has been open for critical metrics, such as reported Covid-19 cases and deaths. This allowed work and understanding to be developed outside of Government departments and regulatory authorities. This helped challenge policy direction (e.g. herd immunity) and identified critical issues, such as the lack of ventilators.

The availability and ease of access to data in for the English NHS have been positive for society. Still, much more could and should be done to (i) develop a better understanding of system constraints and capabilities in advance and during, (ii) create and share more open data more frequently.

On (i) – develop a better understanding of system constraints and capabilities

  • It was and remains unclear how much oxygen can be supplied at peak flow in hospitals. This is because most calculate flow potential in theory and not practice, which involves “hose pipes out windows”.
  • It was unclear how many ventilators existed outside of those used for critical care and where they were stored.

On (ii) – create and share more open data more frequently

  • It is still unclear what system the system capability is for developing tests and delivering PPE. Hopefully, there will be a vaccine, but when there is, it is unclear how quickly this will take to manufacture and distribute.
  • Better and more open data to allow crowdsourcing of intelligence – this should not be seen as creating a “risk of misinformation” (endemic in the NHS), mainly if there is a peer community that can comment and reshare

6c. Responding quickly

In March there were far too few ventilators. At the start of April, every hospital looked at the risk of pushing their oxygen systems beyond their capabilities. It would have been easy to say in those weeks that under-investment in hospital infrastructure was a failure that would lead to increased mortality.

There may be some truth in this, but until hospitals are built to significantly higher standards (which will require more investment than might be available for many years, especially if this is across ALL hospitals) responsiveness is the essential area for development.

This means that responsiveness is a crucial ingredient. This should not be the privilege of Government institutions, such as hospitals, but a population effort, including the public and private sector.


Edge Health are a specialist UK healthcare analytics consultancy that use data and insights to improve the delivery of health and care services, so that better outcomes can be delivered more efficiently.