You may recall me writing two weeks ago about B.1.1.7 being in a race vs. our vaccination rate, and that we needed to get about 100 million people vaccinated by late March to stay ahead of the curve - roughly 1 million per day.

Apparently, the CDC has been following (if only).  They released a report that shows that, if we can vaccinate a million a day, we'll be just ahead of the B.1.1.7 wave and it won't be a surge, just a delay in getting the curve under control.  

*Their numbers are almost identical to my assumptions:*  

1 million vaccinations per day, 10-30% already infected, 60 new confirmed cases/100k as a starting point, and initial Rt of 1.1 or 0.9 with distancing restrictions kept constant.  (That last bit may be optimistic, we've shown that the moment cases start to drop, demands for reopening seem to outweigh the science.)  

<hr>


![image.png](https://files.peakd.com/file/peakd-hive/zafrada/dm0IfqV8-image.png)

#### Some of the CDC estimates are a touch on the optimistic side.



The 10-30% previously infected is reasonable (my guess would be 20%), the 0.5% B.1.1.7 is right on, the 60/100k new case rate is close to where we are. But the 50% more infectious is at the low end of current estimates, and their assumptions that the vaccines prevent 95% of transmission (probably more like 80?) and that previous infection gives 100% immunity are on the high side.) Their estimate that the reporting rate is only 25% might inflate the effect of natural immunity, I think an estimate of 30-40% is probably closer to the mark.

The gray band shows the outcomes depending on previous infections (i.e. progress toward herd immunity). These two graphs show what would happen without a vaccine. On the left, it assumes that our restrictions are currently not quite enough to prevent an increase in cases, which seems likely given our tendency toward lockdown fatigue. On the right, it shows what a minor adjustment (reducing 18% of transmission) would do. The range is stunning, it's the difference between another 600k deaths vs. 200k.

The business-as-usual graph on the left would have cases rising into February, then mostly flat as our slowly increasing number of infected people reduces Rt for a while. Then, as B.1.1.7 becomes dominant in late March or early April, a modest climb, and we'd still be at today's infection and death rate going into May.

(The dark purple shows infections from current strains, the light purple is infections from the new strain.)

In both cases, the fight against COVID would run through the summer.


![image.png](https://files.peakd.com/file/peakd-hive/zafrada/hiD1EFYS-image.png)

These are the same graphs, but with a vaccine introduced at 1 million vaccinations per day. They use a pessimistic assumption of immunity 14 days after the second dose, but it looks like immunity happens, by and large, about 7 days after dose 1...so the effect of the vaccine in these models is delayed by about a month from what I consider likely.

(14 days after dose 2 is what the pharma companies picked, before the trials, as the data point from which they planned to measure efficacy. It was a cautious choice intended to give the highest possible efficacy score. The actual data shows immunity hitting much much sooner.)

The graph on the left is what happens if we don't do any additional restrictions, so that COVID cases would continue gradually rising...and then hold those behaviors constant for the duration. It shows a peak in early February, then the effect of natural and vaccine driven immunity starts to push down the infection rate of existing variants. By the time B.1.1.7 takes over, the battle is already decided, and we enter May at half of our current infection rate. The curve would then trend sharply downward as B.1.1.7 can't expand beyond 100%, and the combination of natural and vaccine immunity shuts it down sometime in June.

The graph on the right is the same thing, but with just a bit more effort - like nationwide restrictions on restaurant capacity or closing bars, etc. In that case, cases would plummet to 1/6 the current level by the start of March, before B.1.1.7 becomes a factor, stay at that low level as B.1.1.7 becomes dominant, then drop to almost nothing in June. My guess is these two graphs represent about 150k and 50k additional deaths, respecitvely.