Reasons Why Arizona Might Have a High Breakthrough Case Rate
Last week I had a post about so-called "breakthrough" cases, i.e. people who get COVID-19 despite being fully vaccinated. The post noted that Arizona has a rate more than twice the national average.
I asked an epidemiologist I know, who until recently worked at the Department of Health and Human Services, why that might be. Here are the three answers I got.
Reopening and Mixing
"One possibility is that reopening has increased mixing among the population and our case numbers are rising again, meaning more exposures than in places with lower case rates." This is true to some extent in Arizona.
As reported in my most recent by the numbers post, a small increase in infection rates started in Arizona about a month ago. Probably not coincidentally, this was shortly after the governor issued a reopening order.
But the rise wasn't that big, increasing only about 2 cases per 100K over the period. Also Arizona is in the lowest tier of states for infection rates. So although there has been reopening and mixing, it doesn't seem like enough to explain a 2x higher rate of breakthrough cases in Arizona.
High Rates of Spike Gene Mutations
A second possibility is variants. Certain mutations of genes responsible for the spike protein on the virus may increase spread and reduce the effectiveness of vaccines. Arizona even has a variant of it's very own.
The CDC publishes surveillance data on the most worrisome variants. They only report data from 26 states because those are the ones with enough variant testing to make numbers meaningful.
They report data on the percent of tests showing the UK (P.1.1.7), Brazil (P.1), South Africa (P.1.351), and California (B.1.427/429) variants. The UK strain is not thought to decrease effectiveness of vaccines or antibodies from infections very much, but the other three are.
Among the 26 states, Arizona ranks 26th in the UK variant (14.1% of tests), 10th in the Brazil variant (0.7%), but 2nd in the California variant (36%). No data is reported for the South Africa variant (it's not clear whether this equals no cases).
So it may be that the California variant is the culprit in our high breakthrough rate. If so then California should have a high breakthrough rate as well, because 53.8% of their tests show the California variant.
California is reporting 2300 breakthrough cases. The JHU dashboard says they have fully vaccinated 8,842,883 people. That works out to 0.026% of fully vaccinated people having breakthrough cases, or a chance of 1 in 3846. That compares to 0.017%, or a chance of 1 in 5,882, in Arizona (see earlier post).
So that lines up pretty well with the difference in variants between the states and means the California variant may be behind our high breakthrough rate.
Seek and Ye Shall Find
This brings us to a third possible explanation. The epidemiologist tells me that:
Breakthrough rates are madding to get a handle on if they're asymptomatic. Identifying asymptomatic infections outside a sampling program is unlikely, so rates of identification vary widely depending on surveillance systems, data aggregation (a national approach vs. not), and sampling programs.
So maybe Arizona is finding more cases because it's looking for them harder.
As noted in an earlier post, random testing is rare. But the three biggest universities in our state have random testing programs. ASU and UA (I don't know about NAU) are also collecting information about who has been vaccinated or not, a crucial bit of info for detecting a breakthrough case.
Phoenix and Tucson also happen to be in the largest population centers. So perhaps we are doing more random testing coupled with collecting vaccination info, in our most populated areas, compared to other states.
The ASU Biodesign dashboard is now reporting spike protein variants detected in their random testing program. Here is the data from the last two months (I computed a 7-day trailing average from their data):
