How long could UK cases continue to rise? And how might hospitalisations increase alongside? A thread... 1/
Despite relatively high vaccination rates compared to other countries, cases are growing and in many areas R is now above 1.5. Remember, immunity is already 'priced in' to this number - without vaccination and the social distancing still in place, R would be *much* higher. 2/
If R is 1.5 and contacts/control remain the same, then we'd need remaining part of the population who could potentially spread COVID to shrink by at least 33% before R drops below 1 & epidemic peaks. This would require additional immunity, either from infections or vaccines. 3/
If contacts were to increase and, say, R increased to 2.5, then we'd need at least 60% reduction in size of remaining group who could potentially spread infection. Again, these calculations already account for vaccination to date – and still rather far from getting R below 1. 4/
Vaccines are rolling out at fairly steady pace but epidemic increasing exponentially, which means infections will gradually outpace vaccines as source of new immunity. The question, of course, is what the impact will be before R drops below 1 and epidemic declines... 5/
Vaccines can directly reduce: the risk of infection following exposure, risk of symptoms following infection and risk of severe disease following symptoms (and can also reduce initial exposure risk in community thanks to herd immunity effects described above). 6/
If cases rising, it's symptoms->severe disease risk we're interested in. COVID admission are rising, and were infected when cases at low levels. So even if other links weakened by vaccines (less exposure, infection, symptoms), cases are translating into hospitalisations... 7/
So hospitalisations will continue to increase unless either we see a big further change in symptoms->severe disease risk (which is unlikely over such a short period), or R eventually declines below 1 (as described above). 8/
Finally, these are all calculations for UK, which has relatively high vaccination rates. The impact of Delta could be much, much worse in other countries – emphasising importance of getting vaccines to people at risk in rest of the world as soon as possible. 9/9
• • •
Missing some Tweet in this thread? You can try to
force a refresh
A common Q: “how can COVID hospitalisations in UK still grow if vaccine % high?” Answer: look at the data. Average was ~120 daily COVID admissions over past week. These would have been infected about 2-3 weeks earlier, when case numbers ~2000 per day. This shows two things... 1/
First, there was still a group at risk of hospitalisation a few weeks ago. And second, this risk was large enough to show up as hundred of admissions in recent data, even though cases were at relatively low levels. 2/
So the key question here: if case numbers were to grow X times larger, why wouldn’t hospitalisations also grow X times larger? 3/
One thing that has hugely shaped countries' response to COVID, and which I don't think gets enough discussion: genuine constraints and perceived constraints. A thread... 1/
Some apparent local constraints have persisted throughout the pandemic. For example if you look globally, there are still notable differences in approaches to surveillance and quarantine... 2/
It's worth reading these papers on the response in Taiwan (jamanetwork.com/journals/jama/…) and Korea (jamanetwork.com/journals/jama/…). Should more countries include these data-intense approaches in future pandemic plans? Or does reluctance to date reflect an immovable constraint? 3/
Preliminary UK data on vaccine effectiveness against B.1.617.2 (originally detected in India) now available: khub.net/documents/1359…. A few things to note... 1/
First and foremost, it’s another reminder that *second doses matter*. By Aug/Sep, UK will be in much better position against B.1.617.2, but there’s a risk of substantial transmission in meantime as things reopen. 2/
Also remember that when vaccine effectiveness high, small absolute differences can have big effect. E.g. a drop from 95% to 90% would double number at risk (and probably more than double outbreak size given non-linear nature of transmission). 3/
In real-time, epidemic data streams are patchy, delayed, biased and often contradictory. That's why scientists use terms like 'realistic possibility', 'medium confidence' etc. Uncertainty is inevitable (although will reduce over time) - and yet decisions still need to be made. 1/
B.1.617.2 has been spreading fast in some areas, and people are working hard to disentangle causes & quantify exactly what it means for wider transmission. Control measures are now both going in (e.g. testing, vaccination) & being relaxed, which making analysis even trickier. 2/
Have recently been thinking more about the transmissibility of endemic seasonal coronaviruses... antibody positivity increases sharply at a relatively low age, suggesting high transmissibility in susceptibile populations (below from: nature.com/articles/s4146…). 1/
Older groups have built immunity to seasonal CoVs, but above suggests R0 (i.e. R in fully susceptible population) could be quite high for these viruses. Possibly so high that even stringent measures wouldn't be enough to control in susceptible pop?
One subtlety of below issue that’s worth highlighting – targeted travel bans (as opposed to near-total border closures) have played out in much the way we’d have expected pre-COVID, delaying rather than stopping local epidemics. A few thoughts… 1/
As noted by @firefoxx66 at the time, targeted bans can delay introductions, but this will be of limited use if measures aren’t also in place to deal with (undetected) local circulation: