Dr Raghib Ali is an Honorary Consultant in Acute Medicine at the Oxford University Hospitals NHS Trust, and a Visiting Research Fellow of the Department of Population Health, University of Oxford.

Last month, I described on this site what I expected the likely trajectory of the second wave to be; explained why a certain level of suppression was needed to enable the NHS to keep running all services running to prevent non-COVID health harms; and why that should not be achieved through more lockdowns, since the Government’s own cost benefit analysis showed it caused greater overall long-term health harm.

Since then, there have been increasingly pubic divisions between both scientists and politicians as to the best way forward over the coming months with opposing declarations and memorandums.  Some say current restrictions go too far, others not far enough – with the Government left in an almost impossible position of choosing the least worst option.

Today, I will briefly review the main strategies that have been proposed – herd immunity/focused protection (Great Barrington Declaration, GBD); further suppression/test & trace (John Snow Memorandum, JSM); the current tier system with targeted restrictions; and a potential alternative way forward.

In my last article, I outlined three criteria that any strategy / intervention should be judged by:

i) the evidence for effectiveness,

ii) whether a mandatory approach produces better outcomes than a voluntary one, and

iii) most importantly, that they produce less overall harm.

And today, I add a fourth – compliance – because all measures are only effective if a high enough proportion of people comply with them.

The Great Barrington Declaration

Having consistently highlighted the health harms of lockdown since May, I did of course welcome the GBD’s emphasis on the many harms of lockdown – particularly in developing countries without welfare states – where lockdowns are even more likely to cause overall health harm.

However, a declaration is not a policy and there are still too many unanswered / unanswerable questions which have been highlighted by many others.

Although I fully agree on the need to focus protection on care homes & hospitals, I am not yet convinced that it is feasible to shield the very large numbers of vulnerable people in the community while Covid-19 transmission is high (especially for those who live in multigenerational households).

Also, the number of ‘non-vulnerable’ who would be symptomatic and hospitalised in the coming three to six months (about 50 per cent of COVID admissions are currently aged 18- 64) would make it very difficult to maintain all NHS services – there is simply not enough spare capacity (particularly of staff.) And no country has successfully followed a herd immunity strategy.

It also does not currently have public support (more than 2:1 oppose it) which would be essential in maintaining compliance with shielding for the vulnerable; and in persuading the less-vulnerable to be exposed by returning to normal life.

Finally, its advocates have not shown that it will cause less overall harm which is, of course, the key overall metric it should be judged by.

The John Snow memorandum

In response to the GBD, the JSM was released. I understand the rationale of those who advocate a second (so-called ‘circuit-breaker’) national lockdown who believe that is better to have a shorter lockdown now than a longer one later – which may well be true – but this is not the key question, which should be: is it less harmful than not having one at all?

The limitations and harms of lockdowns have been well documented and I will only add a few points.

There is insufficient evidence that a two week lockdown will achieve its aims (Israel’s second lockdown has already lasted four weeks) and it may not be possible to lift it after two weeks if cases are still rising. And this strategy may just lead to a cycle of lockdowns which is not sustainable.

Lockdowns are only effective if they are complied with (Israel had much lowers levels of compliance in their second lockdown) and, even with current restrictions in the UK, compliance is lower than it was during lockdown and there is no guarantee it will be high enough to be effective.

The models only show that a lockdown may reduce Covid-19 deaths, but these have not modelled the number of non-Covid lives that will be lost or adverse health effects from other causes.

Although the intention of the circuit breaker is to buy time to get Test &Trace (T&T) back on track and ensure the NHS is prepared, it is hard to see how two weeks would make much difference when the NHS has had months to prepare.

T&T is of course an essential part of the solution but, again, there is insufficient evidence that we will ever be able to control the virus through T&T – we tried this over the summer when virus levels were almost zero after one of the longest lockdowns in Europe and it hasn’t worked here – or in the majority of countries in Europe.

It is also not true to say that ‘the only thing that works is lockdowns’ – social distancing and self-isolation before lockdown here was bringing R down, and Sweden has showed it is possible to overcome a first wave without one – which I will return to later.

Finally, although public support for a two week lockdown is currently high, this would change if it was made clear that it could be 4 or 6 weeks or that it may well cause more long term health harm than benefit.

I know and respect many of the scientists supporting both positions and know they genuinely believe their strategy will cause the least overall harm, but I have not signed either the GBD or JSM. Neither adequately acknowledges the limitations and harms of their approaches or the uncertainties of the evidence – and both are overly confident in their assessment of their effectiveness.

The Government’s three-tier system

I still think the current Government strategy of suppression to keep cases low enough to maintain all NHS services and minimise non-Covid health harms while trying to protect education and jobs is a reasonable compromise. Furthermore, if virus levels get too high, fear increases and people don’t come to hospital, don’t go out and the economy suffers, etc.

I certainly support the targeting restrictions based on the local level of cases as opposed to blanket national ones. I find it hard to understand how it can be possibly be fairer to destroy jobs and businesses all over the country including in areas where hospitalisations are extremely low than to target restrictions on those areas where they are highest and the NHS is under pressure. This should not be a political issue, or North vs. South – it’s just common sense.

We can only get through this crisis by supporting each other, and by keeping the economy open in as many places as possible, we can help fund businesses and jobs in those areas that are forced to temporarily close until the pressure on the NHS subsides.

There is also evidence that the current measures are working – R is stabilising (or even falling) in most regions at about half the level of the first wave, and the NHS is not being overwhelmed. However, they have not yet bought R down below one, and hospitalisations and deaths are still  rising.

The key problem appears to be compliance, and we need to focus more on how we can improve compliance with existing restrictions rather than increasing restrictions. After all, the purpose of restrictions (& lockdowns) is purely to enforce social distancing. We urgently need to analyse levels of compliance by local area and to understand what is driving lack of compliance.

For example, despite good intentions, only 20 per cent of those required to self-isolate are doing so, and it may be that financial incentives may be more effective in sustaining compliance with testing and self-isolation (e.g. paying people to self-isolate, as in Germany and Sweden).

Some people also think the measures aren’t working and so say, ‘What’s the point?”, but this is not true, which needs to be stressed.

So I think it is reasonable for the Government to maintain its current strategy for three more weeks to see the effectiveness of the Tier Two and Three restrictions and devolved nations ‘circuit-breakers’.

A potential Plan B

However, we also need a Plan B to get us to Spring – and potentially longer if vaccines / treatments / mass testing are not as effective as hoped.

I would therefore ask the Government to consider an alternative strategy which may cause less overall harm based on the Swedish approach, but with much better protection of the vulnerable, especially in care homes. This is now much more achievable than in the first wave – because cases are lower, testing and PPE are more available. Individual risk calculators will now also  enable, smarter, voluntary shielding of the community vulnerable.

The key point is that Sweden has shown that it is possible to suppress the virus and get over their first wave (and so far control their second wave) without a national lockdown; without reaching herd immunity and without an effective T&T system. And while keeping schools and businesses open – and so reducing overall harms.

The Swedish approach has been widely misunderstood – the official government strategy is ‘to limit the spread of infection in the country and by doing so, to relieve pressure on the health care system and protect people’s lives, health and jobs.’

And as its Chief Epidemiologist, Dr Anders Tegnell, has said, Sweden is not trying to reach herd immunity (and has not achieved it) and it did not encourage the non-vulnerable to return to normal life. Indeed, its government is strongly encouraged social distancing, reducing social contacts, plus the use of public transport and working from home).

It has also introduced many other measures (i.e: closing universities, table-service only in restaurants, limited gatherings to 50 people).Tegnell has described the policy as a ‘voluntary lockdown’ – and generally the levels of compliance have been very high.

The Government has stressed personal responsibility and trusting the public with simple, consistent, public health messaging and tried to build public consensus and trust.

Of course, there are differences between Sweden and the UK, and there is no guarantee that its approach would work here, but the principles are still valid. The Sweden model is also not a cost-free option – and may lead to more Covid-19 deaths in the short term than would otherwise have been the case – but that is not the key metric, which is whether the strategy will lead to the least overall health harm in the long term.

The key to any successful strategy is sustainable compliance – and it must therefore have public trust and confidence. Open debate is important but, as I wrote in June, ongoing divisions lead to both fear and complacency, undermine public confidence and compliance – and can cost lives and livelihoods.

We therefore need doctors, scientists, and politicians to get behind the same overall strategy and I think this will only be possible if we can show which one causes the least overall harm.

The Government should therefore immediately bring together doctors, scientists and economists to conduct a comprehensive cost-benefit analysis of these four options (similar to the one they have already conducted) and come to a consensus – which should then be shared with the public and other scientists.

I have made my own assessment, but I neither have access to all the data nor a monopoly on wisdom and so am happy to accept whichever option comes out best – and hope others will do the same.

I end with the same conclusion as in June: ‘Finally, of course we are not primarily ‘pro- or anti- lockdowners’ – we are all ‘pro-protecting lives and livelihoods’ and wanting to recover from this crisis as quickly as possible. And so, we must put aside our differences, compromise and come together in the national interest. ‘A house divided against itself cannot stand.’