Thread by @gnomeicide

I've been trying to fit trends to whats happening with Covid-19 data in the UK and I think we may just be seeing some trends emerge. Here's the overall picture - recorded cases and deaths continue to fall (1)

Here's a closer look at the recent trends, on a log scale. You can see that there's a straight line for how fast cases are declining on the log graph, making it a regular proportional drop, whereas deaths are still curving downwards (2)

What does that mean? Well in terms of positive test results it means we saw a fall in positive tests from around the 4th to the 17th of January, and while that fall continues now its no longer accelerating, its more or less a constant 4% fall per day now. Roughly. (3)

Out of interest, I keep telling people that the second wave started when the pubs opened on the 4th of July. Here's what that looks like. Its very clear in the data. (4)

Using recorded infections to calculate R you basically see the same trend. Infections rate stabilised between about 0.75 and 0.85. Which is pretty awful, it isn't falling anything like fast enough but its roughly in line with lockdown 1 (5)

But you'll observe there I've got another line for R there calculated from deaths, but thats at a greater delay. We know when peak infectivity happens, we know how long it takes for Covid to kill (on average) after infection, so we have two ways of back calculating R (6)

We can either use deaths or infections, using the numbers for peak infectivity or deaths. That means R from deaths can only be calculated at a greater delay because it takes longer to kill than to spread infection. The two basically track each other (7)

...well, eventually they tracked each other. If you go all the way back up to tweet (1) you'll see our complete failure to do testing in wave 1 didn't help, we were literally finding out people were infected as they were dying, so very early on they don't track at all. (8)

Since then they kind of do, with changes in testing policy and capacity causing all sorts of noise. Trends track, but its noisy (9)

What does that mean going forward? I'm still nervous to build specific projections around testing data, but we would expect the fastest rate of decline in deaths to slow down now, if deaths test tracking at all. This fast fall may not continue (10)

The trend for deaths followed the trend for infections, the trend for R followed, we should see the rate deaths decline start to level off and produce a straight line on a log plot soon, We may almost be there, looking at this (11)

But this week we'll find out. Today is the first 'proper' day for reporting deaths, one of the days Tues-Sat without a weekend lag, when the numbers are typically horrid. Question is, does this fast decline keep going for now? (12)

If it does we will see the first Tuesday since the 5th of January with under 1000 fatalities. If its slowing down, we won't. The rolling 7 day average is now under 900, which is better than we've seen (it peaked over 1200), we need it to keep declining -fast- (13)

Make no mistake - we're still in a hell of a mess. We are not in a good place. We're averaging 890 deaths a day now, we were averaging 11.5 at this stage in the 1st wave. It is no overstatement to call our handling of this calamitous (14)

From a higher point than last year we have further to go to recover from the second wave. (15)

With new variants of Covid-19 on the horizon leaking like water through our sieve like, non-biosecure borders, and with the efficacy of our vaccines to those being at least questionable, we still face an uphill struggle (16)

With the knowledge vaccines may not suppress R if the South African variant spreads here (which it may ell be doing) it is once again time to ask of our political leaders what it is they'll do to suppress infection FASTER. Not what they'll do to open our economies faster (17)

Those economies with a higher proportion of deaths have suffered more economic damage. Of all major nations, Britain has suffered more deaths per capita at higher economic cost. There is no perspective from which our handling of this has been competent or effective (18)

Vaccines are -part- of the way out, but it is a mathematical certainty that we can't escape this just with vaccines. We need tighter national biosecurity, and we've needed it for over a year. I'm as worried about this now as I have been for a year (fin)