How does one define a Second Wave?

[lljkk]

So many people talking about this... but what does it mean to you - how would you know if we were in a second wave? Conversely, how would you know we had avoided a 2nd wave?

Is any increase in new cases relative to where we are now = 2nd wave?

A second wave = a notable uptick in cases. Literally an uptick - it would show on graphs as the graph turning north again.

It would probably not be visible until a month after it had occurred, minimum, though, as deaths don't occur until at least about 15 days after infection, probably longer, and reporting of deaths adds another couple of weeks minimum.

Of course, if we had a mass testing system then we could potentially spot cases sooner - but we don't, and the many asymptomatic cases will not be picked up in the absence of track and trace, leaving these individuals to infect others.

I'm expecting a second wave to start very soon, with the impact of VE Day and then to continue more steeply, once you take into account Cummings' well-publicised behaviour on compliance rates, and then the ending of the lockdown on Monday.

Obviously time will tell.

I would be overjoyed to be proved wrong. But suspect that that will only happen if the disease has magically mutated to be less aggressive. Because I don't know what else will save us.

So has Italy and Spain avoided a second wave? If so how? Are their restrictions still tougher than ours after lockdown has ended for them or have they got a better track and trace? For a while I seen Italy's death rate mentioned a lot but not seen any numbers for weeks now

Good question, black768.

So if the 7 day rolling avg of new confirmed cases was recently about 2000, then would a move to 2500 be a "notable uptick"? What if that move to 2500 occurred over 6 weeks? How much of an increase is required over how small an amount of time to say "yup that's a second wave" ?

Would you also need to take into account the differences in how 'confirmed cases' are recorded now we have increased testing?

It stands to reason that now more people with mild symptoms are being tested that there would be more cases recorded.

You will only know that you have avoided a second wave once we have herd immunity - either through enough people catching the virus or through a vaccine. Until then there will always be a risk of a second wave.

How do you recognise it: cases going up dramatically and threatening to get out of control. There is going to be a grey area between a bumpy road and a proper second wave. But I'd say, it looks like in the first wave about 5% of the population came into contact with the virus. So before I'd count something as a second wave it would have to affect at least another 2%, say.

Not just an uptick, but one that looks it’s growing and starting to accelerate. It would necessitate the reimposition of stricter rules.

but what is "dramatically" - the 2% ? How quickly? Presumably 2% more over the next year is not a wave. 2% tomorrow is a spike. Where is the line between them?

or "accelerate" -- sustained acceleration, I suppose, but for how long? 1 week? 3 weeks? How long does it need to last?

Funny we all talk about Second Wave - & nobody knows what it is!

I would consider there to be a second wave if we see the number of covid deaths increasing at an exponential rate again.

A second wave will have absolutely nothing to do with Cummings.

It would mean numbers rising again by a couple of hundred a day from where we are now, I would say, and continuing to rise. Anything less would be a bump in the graph caused by easing lockdown and could hopefully be dealt with by locking down locally

For me it would be a further period of exponential growth. Localised outbreaks and small variations in numbers would not be a second wave in my book.

how "small" a variation is not a second wave?

increasing at exponential rate.. for how long? 1 week? 4 weeks?

Sixnearlyseven - do you mean confirmed cases (about 2000/day) or unconfirmed but guessed at (about 8000/day). In the chart I pasted, the difference from 1887 to 2312 is far higher than 200, so if we go back to 2312, we're far into your second wave territory, yes?

hmmm...
my ideas how to define a 2nd wave:

anything that leads to full SAH Order again lasting at least 10 days. Then we are in a 2nd wave of Lockdown.

Sustained exponential growth (on the rolling 7 day average) for at least 10 days (we'd actually need 17 days of data to be sure). Exponential growth = linear when vertical axis is log scale.

I was thinking of daily deaths going up rather than cases, don't think the new cases are showing a second wave just yet.

Daily deaths is not that useful. There is lag in reporting and lag in how long it takes from contracting the virus to death, potentially as long as the best part of a couple of months.

It would show up on the positive testing being done in the community, especially now track and trace has been launched.

A second wave looks unlikely at present. VE day hasn't even blipped after 3 weeks. Relaxations on social activity are making little difference on transmission.

Possibly in the autumn when living outdoors gets less attractive and comfortable, and the weather conditions get better for the virus to survive, but hopefully herd immunity will help to keep it lower than the spread in March.

Countries that have come out of lockdown are not seeing a second wave. It’s not inevitable.

So... second wave is just a "feeling", I guess. That means we'll have lots of arguments in future about whether a 2nd wave has happened, right?

My feeling won't = your feeling won't = anyone else's feeling.

You'll know if it happens. It will not be because one or another person went to a BBQ, it will be massive sustained transmission because the conditions are right for the virus to escalate, usually because the virus has mutated to give itself a better chance; it might mutate to be more or less lethal. It will happen quickly and lead to lockdown, with every other virus the second wave death peak is usually higher than the first. It will not necessarily affect every country the same, just as how infection rates have been variable this time.

.. but what is "massive"? Did we already have "massive sustained" transmission -- do you know this bc of the high death count in UK? else how do you know?

Spain is still in lockdown. The measures are currently fairly similar to the UK’s except for bars and restaurants being open.

We don't know the actual transmission from any country yet until sufficient accurate antibody tests are available. Think back to February / March, the covid tests themselves were unreliable. So far there's a suggestion from about 10,000 tests that it's only ~7% of the country who've had it (17% in London, 5% outside, unclear what testing happened in the other big cities). If true, that leaves 93% who could still get infected. Ipsos Mori might have misjudged the distribution method, maybe far more had it mildly or asymptomatically, nobody knows yet. Spain, Italy, France and Belgium each had figures of around 5% from similar sampling tests. Herd immunity is considered to start at ~60%, so transmission would slow significantly after another 53% of the population has been infected in the UK, 55% in those countries. See how much bigger that is? You won't miss it if it happens. Let's hope it doesn't because antibody sampling methods were wrong, or because it's held back until there's a vaccine, or it mutates ridiculously mildly.

with every other virus the second wave death peak is usually higher than the first

I have only heard this with Spanish flu, do you have any other examples? I didn't think viral outbreaks usually came in multiple waves (except where the virus is endemic and there are localised hotspots).

The measures are currently fairly similar to the UK’s except for bars and restaurants being open

This seems to be a pretty significant difference.