Comment by angelzen | Hacker News Reader

Comment by angelzen | Hacker News Reader

Parent Full thread

Over the summer Florida has been in the news because of an ongoing epidemic spike. Using hospitalization rates, the spike started sometime in early July. By late Sept it is almost gone. The fully vaccinated rate increased from 48% to 59% during the same time period. Is the receding fully explainable by the increase in vaccination rate, or can we consider other forces at play, for example forces of outside human control that have been at play in every single epidemic before the vaccines era?

https://coronavirus.jhu.edu/region/us/florida

https://data.democratandchronicle.com/covid-19-vaccine-track...

Edit: I plotted Rt as a function of eff, assuming r0=7 and vacRate=1. Rt goes below 1 only for values of eff > 0.85, which means that given the numbers for eff you gave in a different post "overall range of 55.6 - 93.6 effectiveness against infecting others", there are plenty of scenarios where vaccines by themselves cannot stop an infection wave even if the population is 100% vaccinated.

https://www.wolframalpha.com/input/?i=plot+y+%3D+7+*+%281+-+...

https://www.wolframalpha.com/input/?i=solve+1+%3D+7+*+%281+-...

Edit2. For a 0.59 vacRate, there are some solutions for R0 * (1 - 0.59 * eff) < 1 with eff in [0, 1] and R0 in [0, 9], specifically there are solutions for R0 < 2.x Not going to cut it for 0.59 vaxx rate and delta, no matter how effective the vaccines are.

https://www.wolframalpha.com/input/?i=plot+1+%3E+y+*+%281+-+...

Some combination of death, natural immunity, vaccination, distancing, mask use, etc. What other forces are you suggesting?

Note that the more populous counties probably increased vaccination at a greater rate, like Miami-Dade went from 56% to 72% over the same time range.

Natural immunity as a direct effect of an infection wave is my first guess. In some circles 'natural immunity' has become a dirty word :/

The assumptions between the equation rt = r0 * (1 - (vacRate * eff)) are too simplistic and unable to explain Florida infection wave behavior. There are more variables that we ignore when focusing only on Rt and vaccines.

The silver lining: At some point the whole population would have been exposed to covid, either through vaccines, infection or both. At that point the whole vaccination rate / effectiveness discussion becomes moot. Perhaps I wish the public discourse would keep an eye on 'estimated population fraction exposed to covid' that includes all plausible factors.

Yes, I brought up that equation merely to illustrate how vaccination alone can impact Rt downward - for a while it was unclear to me exactly how "partial vaccination" (before herd immunity) was a benefit; that helped me see that it really just mean it slows down the doubling rate until you get to Rt=1. But overall, Rt is impacted by all forms of mitigation including natural immunity.

The way I actually use that equation in my personal dashboards is to estimate what Rt "would" be, at today's mitigation levels, if no one had gotten vaccinated. So for instance, Portland's Rt is currently 0.92 (according to one model). By plugging in Portland's vaccination rate and efficacy estimates, you can estimate that Portland's Rt would be around 2 today if not for the vaccinations, if all other mitigation were the same.

On the one hand, 2 is a lot better than those estimates of 5 to 9.5, which means that we're impacted a lot by current mitigation practices (masks, distancing) and natural immunity. On the other hand, 2 is huge! Given estimates on infectious periods, that means that currently our cases halve every 90 days or so, but 2 would mean cases would be doubling every week or so. Gargantuan difference. So just an illustration that vaccination has a big impact and matters a lot.

or maybe they're just reporting data creatively

https://www.reddit.com/r/dataisbeautiful/comments/pxizb1/oc_...

and it seems like it works

Alabama, Georgia, Missisipi, South Carolina all show a similar profile:

https://coronavirus.jhu.edu/region/us/alabama

https://coronavirus.jhu.edu/region/us/georgia

https://coronavirus.jhu.edu/region/us/mississippi

https://coronavirus.jhu.edu/region/us/south-carolina

Edit: Also Oregon and Washington, lest we suspect there is a grand RedState conspiracy at foot.

https://coronavirus.jhu.edu/region/us/oregon

https://coronavirus.jhu.edu/region/us/washington

Reply on news.ycombinator.com