Thread by @DavidSteadson

Right, back to things that matter rather than petty attacks on citizen activists who disagree with the accepted narrative.

There's a pandemic occurring and B117 and other variants on the rampage in Sweden. Why am I optimistic?

First, let's look at this graph.

The outbreak began in Sweden in the beginning of March, and about 3 weeks later the deaths started to mount up, increasing rapidly until mid-April when they began a long, slow decline.

The decline was slower than anywhere else in Europe, but nevertheless a decline.

The question is - why? The obvious answer is that at the end of March, the Swedish government issued various recommendations on physical distancing, working from home etc.

These would all clearly have an effect on R - the reproduction number, a measure of how fast the infection is spreading. Once you get it below 1, the spread is *decreasing*. Now we had very limited testing in Sweden in the Spring so we have little direct measurement of the spread

Mortality however is a good proxy, we know it follows infections by about 3 weeks. So if the peak in mortality was mid-April, then the peak of infections was around the last week of March.

Infections respond to the interventions, but that also takes about a week to show effect

This all adds up, with FHM starting to issue recommendations in mid-March. A week later infections fall, 2-3 weeks after that, mortality falls.

All straightforward and actually quite predictable.

But here's the interesting thing. Look at the graph again. The red lines show how fast mortality was growing. it's about the same in Spring as in Autumn, just happened for longer in the Autumn.

In the Autumn we had virtually the same level of restrictions as in Spring. Some a little easier, some a little harder.

So why did things start to slow much more quickly in Spring than they have now?

I believe it's because of seasonality.

Respiratory diseases like influenza and coronaviruses show what's known as "seasonality". They wax and wane through the year. Influenza, for example, commonly peaks in Sweden around week 6.

Virologists and epidemiologists have argued about why for decades. Lot's of theories have been proposed, for example -

* humidity
* vitamin D levels
* behaviour change
* UV levels

I think it's likely a combination of these things, and other things we haven't thought of.

Some things, like being more social in the spring, might lead to more transmissions, increasing R, other things, like meeting outdoors more than indoors, or riding a bike to work instead of the bus, decrease transmissions.

For an airborne disease, like SARS-CoV-2, and likely influenza, more humidity means virions drop out of the air faster, opening windows means they disperse quicker. More UV may destroy the virus more. D-level affect immune efficiency etc etc

All of these types of things add up in such a way that in Spring risk of infection drops, R goes below 1, and the infections drop away. In Autumn they reverse. People spend more time indoors. They use public transport more. It gets drier, etc etc.

So here's what I think's happened. In Spring an Autumn we've had various NPI (non-pharmaceutical interventions), primarily all about behaviour change, that dropped R significantly - but it didn't drop R below 1. That's why, until recently, cases weren't dropping in the Autumn.

SARS-CoV-2 has an estimated R0, or base reproduction number, between 2 and 3. On average, with nothing "new" being done, each infected person infects between 2 and 3 people.

The NPI in Sweden got it down to about 1.3 or so. Big change!

But it wasn't until the effect of Spring kicked in that it dropped below 1 and the pandemic faded going in to the summer.

Same now in Winter - but in this case I think first the Christmas holidays and then the very cold winter weather led to people being less social, less active

Less active - less transmissions, and cases dropping. Cases dropping, a few weeks later mortality dropping.

Now, the effect of the holidays has passed, and we appear to have R at approximately 1 - neither increasing or decreasing.

So why am I optimistic? It's only about 6 weeks until mid-March, when last year the effect of Spring brought cases right down! If we can just keep things under control until then, we should see the pandemic fade going in to summer, and vaccinations rolling out nationwide.

Ahh ... but what about these new more transmissible variants like B.117 everyone is worried about? Well, let's look at the data. It's currently estimated around 10% of infections in Sweden are already this variant, and it's estimated to be 30-50% more transmissible.

Nationally, we're averaging about 3000 new cases a day. With people being infectious for roughly 5 days, that means right now, around 15000 infectious people. So about 1500 with B.117

So let's see what going to happen with the "B.117 pandemic". With starting cases of ca 1500, incubation time of 5 days, an infection time of 7 days, and an R of 1.5 (50% higher than now) We get a curve like this.

Pink is number of newly infectious people every 2 days, they're the one's testing positive. Of course, not everyone who gets infected gets tested, so the actual number of *cases* is lower.

In about 6 weeks, that means around new 2000 people a day getting infected with B.117.

We're reporting now less than 3000 new cases/day, and that's not all infections. Before Christmas we were reporting *7000* new cases/day!

It seems likely to me that even if B.117 becomes the bulk of infections, we will be having *fewer* cases in 6 wks than we had at Christmas.

And then, if things go like last year, Spring is upon us. In the graph above I've factored in a further 50% decrease in transmission at that time ... and the pandemic fades away, slowly but surely.

It's likely the worst of the pandemic is behind us!

But that's not all! Cases are currently flat, but if we compare changes in case numbers week to week by age, it looks like this.

*Cases are still dropping among the most vulnerable*!!

As a former #longcovid sufferer and ended up in hospital with a reinfection, I can assure you mortality isn't the only thing to worry about with Covid-19 ... but the age data is very promising

On top of that, vaccinations are rolling out in this group, with nearly 300000 receiving at least a first dose.

Even if cases increase, it's likely mortality will continue to drop.

Now, there's a few unknowns in this. We don't what the R will be for B.117 under current Swedish conditions. We don't know what the weather will be like, and we don't know the effect of other variants.

But I do think we're past the worst of this pandemic.

But don't take this as an endorsement of the Swedish strategy. Cases could have been - and still can - be dropped dramatically with a short sharp lockdown. Drawing out the pandemic instead will, like in the Spring, lead to thousands more unnecessary deaths.

We should be wearing masks. We should be limiting schools. We should be staying home as much as possible We can do much to save tens of thousands from being ill and thousands from dying.

But there's definitely a light I can see on the horizon.

(and winter!)