Hypotheses for the geographical distribution of Covid

[johnnychips]

Let me say I have genuine sympathy for any victims of COVID and their families to start with, as what follows might seem a bit abstract.

I have been really fascinated by how the virus is distributed through the UK, and I have an idea that at the moment that the places with the highest rates seem to have no universities themselves, but are close to areas that have. I would suggest that the Medway towns, Haringey, Basildon etc. have higher rates because the virus has spread out of central London, whose rate increased when the students went back in late September. In a similar vein from where I come from, Sheffield saw a huge increase in October, but has now fallen dramatically; yet Doncaster and Rotherham have now much higher rates.

That’s my idea, and I realise it’s full of holes and exceptions: I wonder why Southampton, a university city with its fair share of high-occupancy housing seems to have avoided high rates; Hull has a university, yet cases have only recently shot up; Bristol has seen a similar but not as prominently delayed reaction.

Other factors come into play, such as the level of testing and so on, but I wondered if anybody else had some other explanations for the past and current distribution of COVID in the UK?

 

[Crossover]

I dare say the amount of testing will have something to do with it. As we are Tier 3, another centre has recently popped up at one of the local railway station car parks.

 

[philosopher]

Let me say I have genuine sympathy for any victims of COVID and their families to start with, as what follows might seem a bit abstract.

I have been really fascinated by how the virus is distributed through the UK, and I have an idea that at the moment that the places with the highest rates seem to have no universities themselves, but are close to areas that have. I would suggest that the Medway towns, Haringey, Basildon etc. have higher rates because the virus has spread out of central London, whose rate increased when the students went back in late September. In a similar vein from where I come from, Sheffield saw a huge increase in October, but has now fallen dramatically; yet Doncaster and Rotherham have now much higher rates.

That’s my idea, and I realise it’s full of holes and exceptions: I wonder why Southampton, a university city with its fair share of high-occupancy housing seems to have avoided high rates; Hull has a university, yet cases have only recently shot up; Bristol has seen a similar but not as prominently delayed reaction.

Other factors come into play, such as the level of testing and so on, but I wondered if anybody else had some other explanations for the past and current distribution of COVID in the UK?

The increase in cases over the past weeks seems to be mainly in London and the Southeast which makes wonder if there have been particularly favorable weather conditions in this part of the country for Covid-19 spread. In London over the past few weeks the weather has been generally been quite murky with little wind, so perhaps this type of weather is conducive to Covid-19 spread. Little wind I guess would make improving ventilation less effective at reducing spread.

 

[birchesgreen]

The weather has been the same in the Midlands though and cases have been falling here.

 

[adc82140]

I think the density of housing has a lot to do with it. That's why the west country has lower levels, and the Highlands and islands.

I dare say the amount of testing will have something to do with it. As we are Tier 3, another centre has recently popped up at one of the local railway station car parks.

There were 21,000 positive tests recorded yesterday. However what wasn't mentioned is that over 400,000 tests were processed. I think that's the most ever. Last week when we were clocking in about 12,000 positives, only 230,000 tests were being done.

 

[Yew]

Indeed, for tiering, they seem to be failing to nondimensionalise by the number of tests being carried out, and thus many areas are forever in T3 purgatory

 

[30907]

the density of housing has a lot to do with it. That's why the west country has lower levels, and the Highlands and islands.


There were 21,000 positive tests recorded yesterday. However what wasn't mentioned is that over 400,000 tests were processed. I think that's the most ever. Last week when we were clocking in about 12,000 positives, only 230,000 tests were being done.

Yes, housing density is an obvious factor overall, but it is the variations between similar areas that the OP notes (as have I)

I am no statistician, but one-day figures are not a reliable guide. Using 7-day averages as per HMG et al, there seems to be hardly any correlation between positive tests and number of tests - the latter has fluctuated nationally in the range 200k-400k with an average around 320k, the former much more variable.

Indeed, for tiering, they seem to be failing to nondimensionalise by the number of tests being carried out, and thus many areas are forever in T3 purgatory

My previous comment applies.

On the original theme, a couple of observations:

1. There seems to be a good correlation between the sharp spikes-and-drops in early October and University outbreaks (Liverpool, Nottingham, Manchester) - outbreaks in unis (and well-managed workplaces) being easy to suppress. I make no comments about methods. I can see that localised outbreaks might then have slower over-the-border impacts.

Another example of delayed impact is, apparently, Bradford: I can't locate the report (but it was linked by Dr John Wright whose reports I - and my GP daughter - trust), but the first cases here were not in the inner city but in leafier areas. This suggests that the first infections might be via people who travelled to the hotspot of London - or indeed the Italian ski resorts - and then trickled down to the now-familar areas of relative poverty, crowding and non-WFH jobs.

2. Other early autumn outbreaks across the North were often in areas where the first wave had had relatively low impact - Barnsley for example.