It is three months since we started to leave lockdown and, while we have learnt a lot about the virus in that time, the key questions I discussed then remain. What are the risks of a second wave? And what should the appropriate response to it be?
On the first question, the debate remains as polarised as ever – with two widespread misunderstandings of the data. The first is that cases are now as high as they were during the first wave – leading to unnecessary fear – and the second is that hospitalisations and deaths (and test positivity) are much lower now than when we had a similar number of cases – leading to unjustified complacency.
These case count (and test positivity) comparisons are completely misleading, because far more tests are being carried out now than in the first wave when they were only being done on those is hospital, not in the community.
Therefore, the number of recorded cases were a huge under-estimation – the actual number of cases at that time is now known based on antibody testing. Roughly seven per cent or four million people have had the infection in England, and daily case counts can be inferred based on this, as well as working backwards from the resulting hospitalisations and deaths. .
We are now at roughly 3000 cases a day on the basis of positive tests (the true number is about 6000 based on the ONS survey, with about 40000 cases in the last week). This is about three times higher than the trough in July but less than five per cent of the peak in March (when there were about a million cases a week).
So while there is certainly no need to panic based on case count, those who are only looking at hospitalisations and deaths are also in danger of being falsely reassured. At the beginning of the first wave, it took less than three weeks to go from 6,000 cases a day on March 5th (50 recorded cases) to 300,000 cases a day on March 23rd. (2000 recorded cases). There was only one death on March 3rd, but a month later it was a thousand a day (with 3000 admissions).
The key point is that the situation can change very rapidly and there is no room for complacency. By the time hospital admissions and deaths are going up rapidly, it is too late.
We can also learn from other countries’ experience – again, the evidence is mixed and you can come to different conclusions based on which countries you look at.
However, generally it is true that those countries experiencing large second waves after lifting lockdowns are those that had small first waves so more of their population remain susceptible to infection. But it is not true that the second waves are only causing cases and not deaths (for example, Romania, Israel, Morocco have had far more deaths in their (ongoing) second waves than their first.)
The converse is also true (so far) in that countries with large first waves are having smaller second waves, and particularly in cities like New York where about 20 per cent of the population has developed antibodies, there is no second wave. This latter finding was surprising, as it was expected that herd immunity would develop at a minimum of around 50 per cent of the population being infected, but there is now some evidence that many more people are immune as they have developed immunity from other parts of the immune system or from previous infection with common cold coronaviruses, and this is not detected by routine antibody testing.
However, there is also an exception to this experience: Iran – where a large first wave of deaths has been followed by an even larger second wave (despite about 20 per cent of its population being infected in the first wave), and so we can’t be that confident as to what will happen here.
In general, second waves of cases have started about three months after the first wave and Spain is probably the best guide for us having had a similar level of infections, deaths and immunity in their first wave. Their second wave started about a month ago, and hospitalisations and deaths have now also started increasing – but at a much lower rate than in the first wave.
There has also been significant disagreement about how deadly Covid-19 is – with some incorrectly asserting that Covid-19 is no more deadly than flu. We now have much better evidence with actual data (as opposed to modeling) for the infection fatality rate (IFR) – the percentage of those infected who die (not just those who are diagnosed as positive cases) – with most estimates between 0.5 and 1.0 per cent – whereas flu is less than 0.1 per cent.
Further evidence for this comes from New York, where about 20,000 died – an IFR of 0.1% would mean that 20 million had been infected – but its population is only eight million.
The IFR also varies hugely by age (in children and young adults it is almost zero whereas in the over 65s it is above five per cent) and by country, due to differences in the proportion of the elderly, levels of chronic disease and the provision of healthcare.
In the UK, we now have good evidence from death certificates that Covid-19 was the underlying cause of death in about 50,000 people – it’s not that they died just ‘with Covid-19’ – which tallies with my own experience on the front line back in April. And most estimates for the IFR in the UK have been around one per cent, but it may be half that due to undetected infections.
So where are we now? It is clear that our second wave has started (about a month after Spain, as expected) with cases now doubling roughly every seven to eight days (whereas in the first wave it was every three days – with deaths following the same pattern three weeks later.) Hospital admissions have also doubled over the last two weeks.
In the UK as a whole, we are nowhere near herd immunity (we are likely to be at about 15 per cent maximum) and so we are possibly about a third of the way through the epidemic.
It is therefore theoretically possible that another 100,000 people could die from COVID-19 in a second wave in the coming months. Reports from SAGE and the Academy of Medical Sciences also suggested 85,000 and 120,000 respectively in reasonable worst case scenarios.
However, a repeat of the first wave is very unlikely, as cases will increase more slowly due to the measures now in place including social distancing, masks, hand washing; the massively increased community testing (although clearly not enough) which provides local data to target interventions earlier, and allows tracing and isolation of contacts – plus the fact that we have a higher population level of immunity.
Hospitalisations and deaths should also be significantly lower due to the lower age profile of cases (although antibody testing shows that even in the first wave the highest proportions infected were young adults and the lowest were those aged over 65); better shielding of those at highest risk and possibly a lower viral load due to social distancing and masks.
We are also now much better at managing the disease with more effective treatments. Another factor that should reduce overall excess deaths this winter is that the flu season may be much less severe than usual due to coronavirus measures – as has been the case in Australia.
But it is also important to realise that although the NHS wasn’t overwhelmed by Covid-19 cases (in relation to ward and critical care bed use) in the first wave – and this is something I’ve only appreciated recently – NHS services as a whole were overwhelmed.
The only way the NHS was able to cope was by shutting down many essential services which caused suffering and death for thousands – particularly cancer patients – and huge increases in waiting lists. Therefore, it is essential to keep Covid-19 cases / hospitalisations at a much lower level this time, so as to ensure that all essential NHS services keep running – or we again risk thousands of additional deaths.
In conclusion, sadly, the pandemic is far from over and a second wave has the potential to cause very significant direct and indirect health harm. Doing nothing is clearly not an option – but neither should a second lockdown be, as I will explain on this site next week.