The recent Lancet meta-analysis suggests 2m social distancing doubles the risk compared to 1m.
But is that actually a meaningful difference in policy terms?
It depends on the baseline (starting) risk & the infection rate in the community. Let's illustrate.
But is that actually a meaningful difference in policy terms?
It depends on the baseline (starting) risk & the infection rate in the community. Let's illustrate.
Lancet suggest a baseline risk of an infected person passing it on of 10% (1 in 10).
Actual risk will depend on the situation: probably higher if you spend 3 hours dancing with the person in a sweaty club, but lower if sat at an outside cafe table.
Let’s stick with 10% for now.
Actual risk will depend on the situation: probably higher if you spend 3 hours dancing with the person in a sweaty club, but lower if sat at an outside cafe table.
Let’s stick with 10% for now.
Lancet estimate 1m reduces that risk by a lot from 10% to about 1.8%. i.e. about 2 in 100, a reduction of 8 in 100.
2m halves the risk again to 0.9% a much smaller reduction in absolute risk of about 1 in 100.
2m halves the risk again to 0.9% a much smaller reduction in absolute risk of about 1 in 100.
However, ONS estimate community infection rate to be about 0.06% (6 in 10k).
So for a random contact, the baseline risk goes down from 10% to 0.006% (6 in 100k).
1m reduces it to about 0.001% (1 in 100k)
2m reduces the risk to 0.0005%, a marginal effect of just 1 in 200k.
So for a random contact, the baseline risk goes down from 10% to 0.006% (6 in 100k).
1m reduces it to about 0.001% (1 in 100k)
2m reduces the risk to 0.0005%, a marginal effect of just 1 in 200k.
Now 0.06% is the ONS estimated infection rate for 25 May to 7 June.
Given new cases from P1 & P2 tests have continued to fall, the rate (& the risk reduction from 2m) will be lower by now and, on current trends, lower again when pubs open on 4 July.
Given new cases from P1 & P2 tests have continued to fall, the rate (& the risk reduction from 2m) will be lower by now and, on current trends, lower again when pubs open on 4 July.
That’s an illustration of the risk reduction of getting the infection. The reduction in risk of complications or death is much lower again.
(someone might want to check my decimal places: hard to avoid typos when numbers are this small !)
(someone might want to check my decimal places: hard to avoid typos when numbers are this small !)
Of course having more than 1 encounter increases the risk.
On the other hand, in contexts where the baseline risk is already low (passing someone in a shop, sitting outside, playing cricket), the reduction in risk will be much lower.
On the other hand, in contexts where the baseline risk is already low (passing someone in a shop, sitting outside, playing cricket), the reduction in risk will be much lower.
That is the context in which @BorisJohnson needs to make his policy decision about moving from 2m to 1m.
*halves* not *doubles"
